In README, (second line) do not end URL with a period. This is the most common way of getting broken links when people copy trailing punctuation as part of a link.
For the antibody selections, the data for 37.2D still look especially terrible. This seems like an important antibody. There is virtually no correlation between replicates from different libraries. I remain concerned that this is because the antibody concentrations are too high. Could you re-do the experiments reducing the antibody concentrations until you get into a range where at least for the first two points there is a clear increase in neutralization as you increase the antibody concentration? I can't tell if you changed concentrations from last time, but all your points are clearly still in a range where increasing antibody concentration is not altering the fraction neutralized, which could be a problem.
Organize the additional analyses into subheadings like the rest of the table of contents.
For the validation correlation, increase transparency of points a bit so it is possible to tell when there are overlapping points.
For the neutralization correlation, correlate with the escape value, not the log2 fold change. We have a linear but not a slope one correlation, so better to show a unit-less measure like escape.
In README, (second line) do not end URL with a period. This is the most common way of getting broken links when people copy trailing punctuation as part of a link.
For the antibody selections, the data for 37.2D still look especially terrible. This seems like an important antibody. There is virtually no correlation between replicates from different libraries. I remain concerned that this is because the antibody concentrations are too high. Could you re-do the experiments reducing the antibody concentrations until you get into a range where at least for the first two points there is a clear increase in neutralization as you increase the antibody concentration? I can't tell if you changed concentrations from last time, but all your points are clearly still in a range where increasing antibody concentration is not altering the fraction neutralized, which could be a problem.
Organize the
additional analysesinto subheadings like the rest of the table of contents.For the validation correlation, increase transparency of points a bit so it is possible to tell when there are overlapping points.
For the neutralization correlation, correlate with the escape value, not the log2 fold change. We have a linear but not a slope one correlation, so better to show a unit-less measure like escape.